Analyses
Stefano Coretta
10/09/2021
1 Read data
emilianto <- readRDS("data/rds/emilianto.rds")
emilianto_attitude <- readRDS("data/rds/emilianto_attitude.rds") %>%
mutate(
language = factor(language)
)
contrasts(emilianto_attitude$language) <- "contr.sum"
attitudes <- readRDS("data/rds/attitudes.rds")
pa_spaces <- readRDS("data/rds/pa_spaces.rds")
ty_spaces <- readRDS("data/rds/ty_spaces.rds")2 Checks
eea <- emilianto_attitude %>% dplyr::select(dim_1, passive_prop, active_prop)
vif(as.data.frame(eea))## Variables VIF
## 1 dim_1 1.052377
## 2 passive_prop 1.991188
## 3 active_prop 2.010439cor(emilianto_attitude$passive_prop, emilianto_attitude$active_prop, method = "spearman")## [1] 0.6989767eea <- emilianto_attitude %>% dplyr::select(dim_1, spaces_prop)
vif(as.data.frame(eea))## Variables VIF
## 1 dim_1 1.052305
## 2 spaces_prop 1.052305cor(emilianto_attitude$dim_1, emilianto_attitude$spaces_prop, method = "spearman")## [1] 0.2634808Since passive_prop and active_prop have a
somewhat high correlation (about 0.7), we decided to merge active and
passive spaces into a single measure of spaces proportion
spaces_prop.
The correlation between spaces_prop and
dim_1 is 0.26. The VIF is very close to 1. We believe it is
safe to include both variables in the model.
emilianto_attitude %>%
ggplot(aes(comprehend, dim_1, colour = comprehend)) +
geom_violin() +
geom_jitter(width = 0.1, alpha = 0.5) +
facet_grid(~language)
emilianto_attitude %>%
ggplot(aes(speak, dim_1, colour = speak)) +
geom_violin() +
geom_jitter(width = 0.1, alpha = 0.5) +
facet_grid(~language)
3 Modelling
3.1 comprehend
get_prior(
comprehend ~
language * dim_1 * spaces_prop,
data = emilianto_attitude,
family = cumulative()
)## prior class coef group resp dpar
## (flat) b
## (flat) b dim_1
## (flat) b dim_1:spaces_prop
## (flat) b language1
## (flat) b language1:dim_1
## (flat) b language1:dim_1:spaces_prop
## (flat) b language1:spaces_prop
## (flat) b spaces_prop
## student_t(3, 0, 2.5) Intercept
## student_t(3, 0, 2.5) Intercept 1
## student_t(3, 0, 2.5) Intercept 2
## student_t(3, 0, 2.5) Intercept 3
## student_t(3, 0, 2.5) Intercept 4
## nlpar lb ub source
## default
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## default
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)comprehend_priors <- c(
prior(normal(0, 3), class = Intercept),
prior(normal(0, 1), class = b)
)
comprehend_pcheck <- brm(
comprehend ~
language * dim_1 * spaces_prop,
data = emilianto_attitude,
family = cumulative(),
prior = comprehend_priors,
sample_prior = "only",
backend = "cmdstanr",
cores = 4,
file = "./data/cache/comprehend_pcheck"
)
conditional_effects(comprehend_pcheck, "spaces_prop", conditions = make_conditions(emilianto_attitude, "language", "dim_1"), categorical = TRUE)
comprehend_bm <- brm(
comprehend ~
language * dim_1 * spaces_prop,
data = emilianto_attitude,
family = cumulative(),
prior = comprehend_priors,
backend = "cmdstanr",
cores = 4,
threads = threading(2),
file = "./data/cache/comprehend_bm"
)comprehend_bm## Family: cumulative
## Links: mu = logit; disc = identity
## Formula: comprehend ~ language * dim_1 * spaces_prop
## Data: emilianto_attitude (Number of observations: 581)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## Intercept[1] -3.97 0.45 -4.90 -3.14 1.00 2907
## Intercept[2] -1.81 0.26 -2.35 -1.31 1.00 3391
## Intercept[3] -0.83 0.24 -1.31 -0.38 1.00 3178
## Intercept[4] 0.95 0.23 0.50 1.39 1.00 3225
## language1 -0.31 0.22 -0.74 0.13 1.00 2814
## dim_1 0.10 0.26 -0.41 0.61 1.00 2269
## spaces_prop 4.47 0.61 3.29 5.67 1.00 3709
## language1:dim_1 0.12 0.26 -0.39 0.62 1.00 2423
## language1:spaces_prop 1.01 0.60 -0.15 2.18 1.00 3002
## dim_1:spaces_prop 0.21 0.68 -1.13 1.57 1.00 2763
## language1:dim_1:spaces_prop 0.13 0.67 -1.19 1.46 1.00 2907
## Tail_ESS
## Intercept[1] 2598
## Intercept[2] 3119
## Intercept[3] 3188
## Intercept[4] 3059
## language1 2747
## dim_1 2749
## spaces_prop 3135
## language1:dim_1 2734
## language1:spaces_prop 2883
## dim_1:spaces_prop 2922
## language1:dim_1:spaces_prop 2491
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## disc 1.00 0.00 1.00 1.00 NA NA NA
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).conditions <- expand.grid(
dim_1 = c(-2.5, -1.5, 0, 1.5),
language = c("Emilian", "Esperanto")
)
rownames(conditions) <- unite(conditions, "cond__", everything()) %>% pull(cond__)
ubm_cond_sp <- conditional_effects(comprehend_bm, effects = "spaces_prop", categorical = TRUE, conditions = conditions)
plot(ubm_cond_sp, facet_args = list(ncol = 4))
ggsave("img/compr-brm-sp.png")conditions <- expand.grid(
spaces_prop = c(0, 0.25, 0.5, 0.75, 1),
language = c("Emilian", "Esperanto")
)
rownames(conditions) <- unite(conditions, "cond__", everything()) %>% pull(cond__)
ubm_cond_dim <- conditional_effects(comprehend_bm, effects = "dim_1", categorical = TRUE, conditions = conditions)
plot(ubm_cond_dim, facet_args = list(ncol = 5))
ggsave("img/compr-brm-dim.png")3.2 Speak
speak_bm <- brm(
speak ~
language * dim_1 * spaces_prop,
data = emilianto_attitude,
family = cumulative(),
# Same priors as comprehend_bm
prior = comprehend_priors,
backend = "cmdstanr",
cores = 4,
threads = threading(2),
file = "./data/cache/speak_bm"
)speak_bm## Family: cumulative
## Links: mu = logit; disc = identity
## Formula: speak ~ language * dim_1 * spaces_prop
## Data: emilianto_attitude (Number of observations: 581)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## Intercept[1] -2.54 0.27 -3.06 -2.02 1.00 3488
## Intercept[2] -0.17 0.22 -0.62 0.26 1.00 3753
## Intercept[3] 0.79 0.22 0.37 1.24 1.00 3450
## Intercept[4] 2.36 0.24 1.91 2.83 1.00 3902
## language1 -1.04 0.21 -1.45 -0.61 1.00 2932
## dim_1 0.15 0.26 -0.35 0.66 1.00 2223
## spaces_prop 5.45 0.59 4.32 6.66 1.00 3846
## language1:dim_1 -0.01 0.26 -0.52 0.52 1.00 2486
## language1:spaces_prop 1.70 0.56 0.58 2.80 1.00 2970
## dim_1:spaces_prop 0.01 0.65 -1.23 1.27 1.00 2546
## language1:dim_1:spaces_prop 0.14 0.66 -1.14 1.46 1.00 2836
## Tail_ESS
## Intercept[1] 2976
## Intercept[2] 2964
## Intercept[3] 3036
## Intercept[4] 3244
## language1 2481
## dim_1 2545
## spaces_prop 2260
## language1:dim_1 2430
## language1:spaces_prop 2895
## dim_1:spaces_prop 2916
## language1:dim_1:spaces_prop 2688
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## disc 1.00 0.00 1.00 1.00 NA NA NA
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).conditions <- expand.grid(
dim_1 = c(-2.5, -1.5, 0, 1.5),
language = c("Emilian", "Esperanto")
)
rownames(conditions) <- unite(conditions, "cond__", everything()) %>% pull(cond__)
sbm_cond_sp <- conditional_effects(speak_bm, effects = "spaces_prop", categorical = TRUE, conditions = conditions)
plot(sbm_cond_sp, facet_args = list(ncol = 4))
ggsave("img/speak-brm-sp.png")conditions <- expand.grid(
spaces_prop = c(0, 0.25, 0.5, 0.75, 1),
language = c("Emilian", "Esperanto")
)
rownames(conditions) <- unite(conditions, "cond__", everything()) %>% pull(cond__)
sbm_cond_dim <- conditional_effects(speak_bm, effects = "dim_1", categorical = TRUE, conditions = conditions)
plot(sbm_cond_dim, facet_args = list(ncol = 5))
ggsave("img/speak-brm-dim.png")3.3 Read and write
rw_bm <- brm(
read_write ~
dim_1 * spaces_prop,
data = emilianto_attitude %>% filter(language == "Emilian"),
family = bernoulli(),
# Same priors as comprehend_bm
prior = comprehend_priors,
backend = "cmdstanr",
cores = 4,
threads = threading(2),
file = "./data/cache/rw_bm"
)rw_bm## Family: bernoulli
## Links: mu = logit
## Formula: read_write ~ dim_1 * spaces_prop
## Data: emilianto_attitude %>% filter(language == "Emilian (Number of observations: 431)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept -1.95 0.18 -2.29 -1.60 1.00 3290 2411
## dim_1 0.30 0.22 -0.14 0.73 1.00 2563 2703
## spaces_prop 7.21 0.66 5.92 8.50 1.00 3595 2820
## dim_1:spaces_prop -0.08 0.75 -1.52 1.41 1.00 2569 2719
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).conditional_effects(rw_bm, effects = "spaces_prop:dim_1")
ggsave("img/rw-brm-spaces.png")
conditional_effects(rw_bm, effects = "dim_1:spaces_prop")
3.4 Age
comprehend_bm_age <- brm(
comprehend ~
language * age,
data = emilianto_attitude,
family = cumulative(),
# Same priors as comprehend_bm
prior = comprehend_priors,
backend = "cmdstanr",
cores = 4,
threads = threading(2),
file = "./data/cache/comprehend_bm_age"
)conditions <- expand.grid(
language = c("Emilian", "Esperanto")
)
rownames(conditions) <- unite(conditions, "cond__", everything()) %>% pull(cond__)
ubm_cond_age <- conditional_effects(comprehend_bm_age, effects = "age", categorical = TRUE, conditions = conditions)
plot(ubm_cond_age, facet_args = list(ncol = 5))
speak_bm_age <- brm(
speak ~
language * age,
data = emilianto_attitude,
family = cumulative(),
# Same priors as comprehend_bm
prior = comprehend_priors,
backend = "cmdstanr",
cores = 4,
threads = threading(2),
file = "./data/cache/speak_bm_age"
)conditions <- expand.grid(
language = c("Emilian", "Esperanto")
)
rownames(conditions) <- unite(conditions, "cond__", everything()) %>% pull(cond__)
sbm_cond_age <- conditional_effects(speak_bm_age, effects = "age", categorical = TRUE, conditions = conditions)
plot(sbm_cond_age, facet_args = list(ncol = 5))
3.5 Mean competence
emilianto_attitude %>%
ggplot(aes(mean_comp)) +
geom_density() +
geom_rug()
emilianto_attitude %>%
ggplot(aes(language, mean_comp)) +
geom_jitter(width = 0.2, alpha = 0.5, height = 0) +
expand_limits(y = 0)
emilianto_attitude %>%
ggplot(aes(language, mean_comp, colour = dim_1)) +
geom_jitter(width = 0.2, alpha = 0.5, height = 0) +
expand_limits(y = 0)
emilianto_attitude %>%
ggplot(aes(dim_1, mean_comp)) +
geom_point() +
facet_grid(~ language)
mean_comp_bm <- brm(
mean_comp ~
language *
dim_1,
data = emilianto_attitude,
family = zero_one_inflated_beta(),
backend = "cmdstanr",
cores = 4,
file = "./data/cache/mean_comp_bm"
)
mean_comp_bm## Family: zero_one_inflated_beta
## Links: mu = logit; phi = identity; zoi = identity; coi = identity
## Formula: mean_comp ~ language * dim_1
## Data: emilianto_attitude (Number of observations: 578)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## Intercept 0.15 0.03 0.09 0.19 1.00 5663 3298
## language1 -0.20 0.03 -0.25 -0.15 1.00 5236 3257
## dim_1 0.08 0.04 0.02 0.15 1.00 3827 2833
## language1:dim_1 0.03 0.04 -0.04 0.10 1.00 4044 3466
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## phi 14.48 0.83 12.93 16.14 1.00 5803 3109
## zoi 0.00 0.00 0.00 0.01 1.00 4083 2039
## coi 0.50 0.28 0.03 0.98 1.00 5738 2396
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).conditional_effects(mean_comp_bm, effects = "dim_1:language")
3.6 Rural vs urban
3.6.1 comprehend
emilian <- emilianto %>%
filter(language == "Emilian")emil_rur <- emilian %>%
mutate(
ru_ur = ifelse(
str_detect(birth_place, "-RU"), "rural",
ifelse(
str_detect(birth_place, "-UR"), "urban",
NA
)
)
)
emil_rur_clean <- emil_rur %>%
dplyr::select(id, comprehend, speak, read_write, educated:familiar, ru_ur) %>%
mutate(
comprehend = ordered(comprehend, levels = c("NO", "AL", "50/50", "G", "VG")),
speak = ordered(speak, levels = c("NO", "AL", "50/50", "G", "VG")),
across(educated:familiar, ~ as.ordered(.x))
) %>%
drop_na()comprehend_rur <- emil_rur_clean %>%
count(comprehend, ru_ur) %>%
group_by(ru_ur) %>%
mutate(
ru_ur_prop = n / sum(n)
) %>%
ungroup() %>%
mutate(
comprehend = factor(comprehend, ordered = FALSE),
ru_ur = factor(ru_ur, levels = c("rural", "urban"))
)get_prior(
n ~
comprehend * ru_ur,
data = comprehend_rur,
family = Beta()
)## prior class coef group resp dpar
## (flat) b
## (flat) b comprehend50D50
## (flat) b comprehend50D50:ru_ururban
## (flat) b comprehendAL
## (flat) b comprehendAL:ru_ururban
## (flat) b comprehendG
## (flat) b comprehendG:ru_ururban
## (flat) b comprehendVG
## (flat) b comprehendVG:ru_ururban
## (flat) b ru_ururban
## student_t(3, 0, 2.5) Intercept
## gamma(0.01, 0.01) phi
## nlpar lb ub source
## default
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## (vectorized)
## default
## 0 defaultcomprehend_rur_priors <- c(
prior(normal(0, 3), class = Intercept),
prior(normal(0, 1), class = b)
)
comprehend_rur_bm <- brm(
ru_ur_prop ~
comprehend * ru_ur,
data = comprehend_rur,
family = Beta(),
prior = comprehend_rur_priors,
backend = "cmdstanr",
cores = 4,
threads = threading(2),
file = "./data/cache/comprehend_rur_bm"
)
comprehend_rur_bm## Family: beta
## Links: mu = logit; phi = identity
## Formula: ru_ur_prop ~ comprehend * ru_ur
## Data: comprehend_rur (Number of observations: 9)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## Intercept -1.84 0.60 -2.94 -0.63 1.00 2740
## comprehendAL -0.47 0.64 -1.72 0.76 1.00 3494
## comprehend50D50 0.07 0.65 -1.22 1.30 1.00 3394
## comprehendG 1.09 0.71 -0.41 2.39 1.00 2570
## comprehendVG 1.40 0.73 -0.23 2.71 1.00 2429
## ru_ururban -0.55 0.56 -1.58 0.61 1.00 2899
## comprehendAL:ru_ururban 0.35 0.81 -1.32 1.84 1.00 2927
## comprehend50D50:ru_ururban 0.12 0.77 -1.44 1.57 1.00 3467
## comprehendG:ru_ururban 0.48 0.73 -1.01 1.85 1.00 3506
## comprehendVG:ru_ururban 0.60 0.72 -0.86 1.98 1.00 3523
## Tail_ESS
## Intercept 2060
## comprehendAL 3066
## comprehend50D50 3127
## comprehendG 2331
## comprehendVG 2556
## ru_ururban 2931
## comprehendAL:ru_ururban 2724
## comprehend50D50:ru_ururban 2861
## comprehendG:ru_ururban 2901
## comprehendVG:ru_ururban 2648
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## phi 15.62 16.42 2.21 58.91 1.00 1528 1461
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).conditional_effects(comprehend_rur_bm, "comprehend:ru_ur")
3.6.2 Speak
speak_rur <- emil_rur_clean %>%
count(speak, ru_ur) %>%
group_by(ru_ur) %>%
mutate(
ru_ur_prop = n / sum(n)
) %>%
ungroup() %>%
mutate(
speak = factor(speak, ordered = FALSE),
ru_ur = factor(ru_ur, levels = c("rural", "urban"))
)speak_rur_bm <- brm(
ru_ur_prop ~
speak * ru_ur,
data = speak_rur,
family = Beta(),
prior = comprehend_rur_priors,
backend = "cmdstanr",
cores = 4,
file = "./data/cache/speak_rur_bm"
)
speak_rur_bm## Family: beta
## Links: mu = logit; phi = identity
## Formula: ru_ur_prop ~ speak * ru_ur
## Data: speak_rur (Number of observations: 10)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## Intercept -1.98 0.35 -2.56 -1.19 1.00 1406
## speakAL 0.83 0.46 -0.18 1.61 1.00 1586
## speak50D50 0.11 0.46 -0.91 0.90 1.00 2493
## speakG 1.05 0.48 -0.10 1.83 1.00 1255
## speakVG 0.61 0.47 -0.49 1.38 1.00 1513
## ru_ururban -0.24 0.38 -0.98 0.53 1.00 2305
## speakAL:ru_ururban 0.56 0.54 -0.63 1.59 1.00 2489
## speak50D50:ru_ururban 0.54 0.59 -0.72 1.62 1.00 2373
## speakG:ru_ururban -0.00 0.54 -1.03 1.13 1.00 2765
## speakVG:ru_ururban 0.02 0.56 -1.11 1.13 1.00 2624
## Tail_ESS
## Intercept 1567
## speakAL 2136
## speak50D50 2401
## speakG 1808
## speakVG 2227
## ru_ururban 2590
## speakAL:ru_ururban 2098
## speak50D50:ru_ururban 2599
## speakG:ru_ururban 2772
## speakVG:ru_ururban 2435
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## phi 73.50 70.06 8.03 268.36 1.00 486 1114
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).conditional_effects(speak_rur_bm, "speak:ru_ur")
3.6.3 Read and write
rw_rur <- emil_rur_clean %>%
count(read_write, ru_ur) %>%
group_by(ru_ur) %>%
mutate(
ru_ur_prop = n / sum(n)
) %>%
ungroup() %>%
mutate(
ru_ur = factor(ru_ur, levels = c("rural", "urban"))
)rw_rur_bm <- brm(
ru_ur_prop ~
read_write * ru_ur,
data = rw_rur,
family = Beta(),
prior = comprehend_rur_priors,
backend = "cmdstanr",
cores = 4,
file = "./data/cache/rw_rur_bm"
)
rw_rur_bm## Family: beta
## Links: mu = logit; phi = identity
## Formula: ru_ur_prop ~ read_write * ru_ur
## Data: rw_rur (Number of observations: 4)
## Draws: 4 chains, each with iter = 2000; warmup = 1000; thin = 1;
## total post-warmup draws = 4000
##
## Population-Level Effects:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS
## Intercept 0.49 0.50 -0.69 1.33 1.00 1758
## read_write -1.01 0.64 -2.03 0.50 1.00 1311
## ru_ururban -0.08 0.54 -1.11 1.14 1.00 1848
## read_write:ru_ururban 0.18 0.67 -1.29 1.35 1.00 1415
## Tail_ESS
## Intercept 1696
## read_write 1930
## ru_ururban 2066
## read_write:ru_ururban 2175
##
## Family Specific Parameters:
## Estimate Est.Error l-95% CI u-95% CI Rhat Bulk_ESS Tail_ESS
## phi 40.90 53.09 1.71 204.55 1.01 664 727
##
## Draws were sampled using sample(hmc). For each parameter, Bulk_ESS
## and Tail_ESS are effective sample size measures, and Rhat is the potential
## scale reduction factor on split chains (at convergence, Rhat = 1).conditional_effects(rw_rur_bm, "read_write:ru_ur")
3.7 SEM
emilianto_attitude <- emilianto_attitude %>%
mutate(
dim_1_z = as.vector(scale(dim_1)),
spaces_prop_z = as.vector(scale(spaces_prop)),
speak_ot = speak
)
contrasts(emilianto_attitude$speak_ot) <- "contr.treatment"bf_1 <- bf(speak ~ language * dim_1_z * spaces_prop_z) + cumulative()
bf_2 <- bf(comprehend ~ language * dim_1_z * spaces_prop_z) + cumulative()
bf_3 <- bf(dim_1_z ~ language * spaces_prop_z * (speak + comprehend)) + gaussian()
bf_4 <- bf(spaces_prop_z ~ language * dim_1_z * (speak + comprehend)) + zero_inflated_beta()
sem_1 <- brm(
bf_1 + bf_2 + bf_3 + bf_4,
data = emilianto_attitude,
cores = 4,
threads = threading(2),
backend = "cmdstanr"
)bf_1 <- bf(speak_ot ~ language * dim_1_z * spaces_prop_z) + cumulative()
bf_3 <- bf(dim_1_z ~ language * spaces_prop_z * speak_ot) + gaussian()
bf_4 <- bf(spaces_prop_z ~ language * dim_1_z * speak_ot) + gaussian()
sem_2_prior <- c(
prior(normal(0, 10), class = Intercept),
prior(normal(0, 10), class = b),
prior(cauchy(0, 1), class = sigma, resp = "spacespropz"),
prior(cauchy(0, 1), class = sigma, resp = "dim1z")
)
sem_2 <- brm(
bf_1 + bf_3 + bf_4,
data = emilianto_attitude,
cores = 4,
threads = threading(2),
backend = "cmdstanr",
file = "data/cache/sem_2",
prior = sem_2_prior
)summary(sem_2)conditions <- expand.grid(
language = c("Emilian", "Esperanto")
)
rownames(conditions) <- unite(conditions, "cond__", everything()) %>% pull(cond__)
sbm_cond_sp <- conditional_effects(sem_2, effects = c("speak_ot", "dim_1_z"), resp = "spacespropz", conditions = conditions)
plot(sbm_cond_sp, facet_args = list(ncol = 4))